2024 — Scientific Reports
Detection of atmospheric radon concentration anomalies and their potential for earthquake forecasting using Random Forest analysis
Random Forest machine learning was applied to atmospheric radon concentration data to detect anomalies potentially associated with earthquakes. The method achieved improved detection accuracy compared to conventional approaches, demonstrating that machine learning can enhance the identification of precursory radon signals for earthquake forecasting research.
2021 — Scientific Reports
Radon degassing triggered by tidal loading before an earthquake
Anomalous atmospheric radon concentration changes before an earthquake were found to be associated with tidal loading. A physical model linking crustal strain from tidal forces to enhanced radon degassing from the ground was proposed, providing a mechanistic link between tidal triggering and precursory radon signals.
2021 — Scientific Reports
Preseismic atmospheric radon anomaly associated with 2018 Northern Osaka earthquake
A significant decrease in atmospheric radon concentration was detected before the 2018 Northern Osaka earthquake (M6.1). Statistical analysis confirmed this anomaly as a precursory signal, supporting the hypothesis that crustal strain changes before earthquakes can modulate radon emanation from the ground surface.
2018 — Scientific Reports
Non-parametric detection of atmospheric radon concentration anomalies related to earthquakes
A non-parametric anomaly detection method based on singular spectrum transformation was applied to atmospheric radon concentration time series. The method successfully identified anomalous radon changes that preceded several earthquakes, providing a statistical framework for earthquake precursor detection without assuming specific signal models.